PHUSE EU Connect 2026

PHUSE EU Connect 2026 will take place in Glasgow for the first time from November 15 to 18. The international conference provides a platform for professionals from life sciences, clinical research, and data science to discuss current developments, share practical experience, and explore new approaches to data-driven processes in the pharmaceutical industry.

The program includes talks, interactive sessions, and numerous opportunities to exchange ideas and connect with fellow professionals. Topics such as AI and agentic AI, open-source technologies, automation, and modern data processes will play a central role.

HMS will once again be there in 2026, sharing insights and experience from real-world projects and applications in the pharmaceutical industry.

Visit the official PHUSE EU Connect website 

Our contribution to PHUSE

Talk

What Is an AI Agent, and How Can It Be Built and Tailored for Pharma Use Cases?

Pratik Khanna

AI agents are the building blocks of agentic AI. They extend the capabilities of AI models by enabling them to reason, solve tasks autonomously, interact with their environment, and use a variety of tools and data sources.

In this beginner-friendly session, Pratik Khanna uses an Exploratory Data Analysis (EDA) agent to demonstrate how more advanced agentic systems can be designed and tailored to specific use cases. The live demo combines data access, cleaning, exploration, and visualization with report generation, code generation, and machine learning in a single interactive workflow. The example illustrates how the underlying agentic design pattern can be adapted to a wide range of pharmaceutical workflows.

Talk

HepCHAT: Accelerating DILI Drug Safety Assessment with Knowledge Graph-Powered AI Agents

Pascal Sanchez (HMS Analytical Software), Cecília Nunes (Merck)

AI agents are the building blocks of agentic AI. They extend the capabilities of AI models by enabling them to reason, solve tasks autonomously, interact with their environment, and use a variety of tools and data sources.

In this beginner-friendly session, Pratik Khanna uses an Exploratory Data Analysis (EDA) agent to demonstrate how more advanced agentic systems can be designed and tailored to specific use cases. The live demo combines data access, cleaning, exploration, and visualization with report generation, code generation, and machine learning in a single interactive workflow. The example illustrates how the underlying agentic design pattern can be adapted to a wide range of pharmaceutical workflows.

Talk

Establishing a Qualified R Package for PK/PD Analysis After SAS Migration

Corinna Stöckinger (Boehringer Ingelheim), Milena Kraus (HMS Analytical Software)

Following a GenAI-assisted migration of legacy SAS code, the project has evolved into pkadsbuildr, a fully R-native package for PK/PD analysis. It provides standardized templates for data transformation pipelines as well as more than 40 domain-specific functions covering areas such as imputation, baseline derivation, date-time handling, and consistency checks.

Beyond its business logic, a complete software development lifecycle has been established, including version control, unit testing, structured documentation, and specifications as the basis for internal R package qualification. Supporting AI tools also help statistical programmers adapt the package templates to study-specific requirements.

The talk provides insights into the package, its qualification process, and how statistical programmers are supported in their day-to-day work.